CUSTOMER STORY · AI CUSTOMER ENGAGEMENT
OneXBots: Building an AI customer-engagement platform on AWS.
Amazon Bedrock · CRM integration · Enterprise AIAT A GLANCE
The project at a glance.
OneXBots provides AI staff for customer support, sales and other business roles. Organizations can train virtual staff on business documents, embed conversational experiences into websites and connect channels such as Zalo and Messenger.
TitanBases designed an AWS platform pattern that separates application runtime, structured CRM data, controlled knowledge assets and managed foundation-model access. The publication package documents POC and production-readiness scope, but does not contain approved operating measurements.
Solution implemented
- Mapped customer channels and embedded experiences into a defined application boundary.
- Separated Amazon Aurora PostgreSQL application data from documents and assets in Amazon S3.
- Established Amazon Bedrock as a managed model-access boundary for controlled iteration.
Delivery outcomes
- End-to-end request handling and CRM-connected context were placed in one reviewable flow.
- Monitoring, access control, failure handling and scale testing remain explicit production-readiness requirements.
- No benchmark, availability, response-time or cost outcome is asserted without approved evidence.
ABOUT ONEXBOTS
AI staff become a platform challenge when conversations connect to knowledge and CRM context.
OneXBots creates AI-assisted experiences for customer support, sales, education and other business roles across embedded web and messaging channels.
As those experiences move beyond isolated demonstrations, the surrounding system must coordinate application state, enterprise knowledge, CRM data and model requests without coupling every responsibility to one component.
AI staff
Role-oriented conversational assistance for business teams.
Enterprise knowledge
Controlled documents support context-rich responses.
Channel integration
Website, Zalo and Messenger entry points.
CRM workflows
Structured customer and application data remain outside the model layer.
TitanBases designed an AWS application and data pattern that separates Amazon EC2, Aurora PostgreSQL, Amazon S3 and Amazon Bedrock. Implementation status and operating KPIs remain subject to as-built evidence review.
THE CHALLENGE
Conversations · Knowledge · CRM context · Model access
The AI response is only one step in a reliable customer journey.
OneXBots needed an AWS foundation that could preserve structured application and CRM data, manage controlled knowledge assets and support growing conversational traffic.
Tightly coupling application, data, knowledge and model services would make prompt, model and channel changes harder to isolate and operate.
The publication package covers AWS architecture and POC/production-readiness validation. Before publication, the delivery owner should confirm each deployed service, production status and permission to use the customer name and logo.
ENGINEERING AT SCALE · AI CUSTOMER ENGAGEMENT
Four operating boundaries around one customer experience.
Customer channels, application processing, structured data and managed model access can evolve independently while remaining part of one end-to-end request path.
The source provides no verified benchmark or SLA. These are architecture and validation boundaries, not production KPI results.
THE SOLUTION
Application · CRM data · Knowledge · Foundation models
Treat AI as one managed layer in the platform, not the entire platform.
Amazon EC2 provides application runtime. Amazon Aurora PostgreSQL stores structured application and CRM-related data, while Amazon S3 holds controlled documents and platform assets.
Amazon Bedrock provides the model integration boundary. That separation gives the engineering team clearer controls for access, monitoring, failure isolation and future model choices.
SOLUTION ARCHITECTURE
AWS Architecture · OneXBots platform pattern
TOPOLOGY FROM THE PUBLICATION PACKAGE · AS-BUILT CONFIRMATION REQUIRED
Customer channels, application, data and model access
Website embeds, Zalo, Messenger and CRM-connected workflows.
Amazon EC2 request handling, prompt preparation and orchestration.
Aurora PostgreSQL and controlled assets in Amazon S3.
Amazon Bedrock as the managed integration boundary.
POC · PRODUCTION READINESS
A clear AWS design for validating a multi-channel AI platform.
These items describe architecture and validation scope from the publication package. They do not assign unverified performance targets.
DESIGNED
EC2
Application boundary for conversations and business processingDESIGNED
Aurora
Structured application and CRM data separated from the modelDESIGNED
S3
Storage boundary for documents and knowledge assetsDESIGNED
Bedrock
Managed integration boundary for foundation modelsImplementation and as-built status, customer-name approval and production KPIs require delivery-owner confirmation. No benchmark, uptime, response-quality or cost result is claimed.
TECHNOLOGIES
A focused AWS stack for application, data, knowledge and AI.
Only services confirmed by the supplied publication package are listed.
APPLICATION
STRUCTURED DATA
KNOWLEDGE
GENERATIVE AI
PRODUCTION · NEXT STEPS
From POC to an enterprise-ready AI platform.
When the delivery owner validates the as-built architecture and implementation evidence, future revisions can add approved performance and response-quality evidence without changing the Golden Visual.
CUSTOMER STORY · TECHNOLOGY AND SAASRELATED CAPABILITIES
RELATED CASE STUDIES
Explore more stories.
RELATED INSIGHTS
engineering insight
Buying ChatGPT Business in Vietnam: A Procurement Checklist for Teams
A practical procurement guide for teams in Vietnam comparing ChatGPT Business Standard and Premium pricing, seat planning, signup steps, privacy controls and onboarding support.
Nov 5, 2026
engineering insight
What Is AI Migration? Can Enterprises Migrate AI Like Cloud Workloads?
AI migration is more than switching model APIs. It means discovering, adapting and validating the behavior, knowledge, tools, memory, permissions and workflows that make an enterprise AI workload work.
Oct 25, 2026
engineering insight
What Is ChatGPT Business? Standard and Premium Pricing, Seats and Signup
A practical guide to ChatGPT Business Standard and Premium pricing, minimum seats, workspace administration, privacy and signup for companies.
Nov 5, 2026
Prepared by TitanBases Engineering Team
TitanBases engineers share practical guidance from cloud, data and AI delivery work, with technical review focused on production architecture, security and operations.
YOUR NEXT WORKLOAD